We consider the determination of a conductivity function in a two-dimensional
domain from the Cauchy data of the solutions of the conductivity equation on the
boundary. We prove uniqueness results for this inverse problem, posed by Calderón,
for conductivities that are degenerate, that is, they may not be bounded from above
or below. Elliptic equations with such coefficient functions are essential for
physical models used in transformation optics and the study of metamaterials,
e.g., the zero permittivity materials. In particular, we show that the elliptic
inverse problems can be solved in a class of conductivities which is larger than
. Also,
we give new counterexamples for the uniqueness of the inverse conductivity problem.
We say that a conductivity is visible if the inverse problem is solvable so that
the conductivity inside of the domain can be uniquely determined, up to
a change of coordinates, using the boundary measurements. The original
counterexamples for the inverse problem are related to the invisibility cloaking.
This means that there are conductivities for which a part of the domain is
shielded from detection via boundary measurements and even the existence of
the shielded domain is hidden. Such conductivities are called invisibility
cloaks.
In the present paper, we identify the borderline of the visible conductivities and
the borderline of invisibility cloaking conductivities. Surprisingly, these borderlines
are not the same. We show that between the visible and the cloaking conductivities,
there are the electric holograms. These are conductivities which create an illusion of a
nonexisting body. Such conductivities give counterexamples for the uniqueness of the
inverse problem which are less degenerate than the previously known ones. These
examples are constructed using transformation optics and the inverse maps of
the optimal blow-up maps. The proofs of the uniqueness results for inverse
problems are based on the complex geometrical optics and the Orlicz space
techniques.